Aquaculture International

, Volume 17, Issue 5, pp 437–448

Mytilus species under rope culture in Scotland: implications for management

Authors

    • Fisheries Research Services Marine Laboratory
    • Institute of Biological and Environmental SciencesUniversity of Aberdeen
  • M. Bland
    • Fisheries Research Services Marine Laboratory
  • A. M. Shanks
    • Fisheries Research Services Marine Laboratory
  • A. Beaumont
    • School of Ocean Sciences, College of Natural SciencesBangor University
  • S. B. Piertney
    • Institute of Biological and Environmental SciencesUniversity of Aberdeen
  • I. M. Davies
    • Fisheries Research Services Marine Laboratory
  • M. Snow
    • Fisheries Research Services Marine Laboratory
Article

DOI: 10.1007/s10499-008-9214-6

Cite this article as:
Dias, P.J., Bland, M., Shanks, A.M. et al. Aquacult Int (2009) 17: 437. doi:10.1007/s10499-008-9214-6

Abstract

Mussel (Mytilus spp.) production is one of the most economically important aspects of global aquaculture and, in Scotland, production has increased from 262 t in 1986 to 4,219 t in 2006. Until recently, mussel production in Scotland was considered to be based exclusively on the blue mussel Mytilus edulis, the native species. In Europe, production is known to consist of M. edulis and M. galloprovincialis, while a third less profitable species M. trossulus occurs in the Baltic Sea, where it is unexploited. In Scotland, the sympatric occurrence of M. edulis, M. trossulus, M. galloprovincialis and their hybrids in cultivation in Loch Etive has recently been reported, with significant losses in production attributed to the presence of M. trossulus. Samples of mussels were taken at three depths from 10 rope-farm sites in Loch Etive. The distribution of the Mytilus species and their hybrids in Loch Etive was investigated using the Me 15/16 nuclear DNA locus for species and hybrid identification. All three species and their hybrids were detected and the data were in agreement with the Hardy–Weinberg model suggesting panmixia. Frequencies of M. galloprovincialis and its hybrids were very low. Overall, M. trossulus (37%) was found to be more common than M. edulis (30%) and 23% of the sampled mussels were M. trossulus × M. edulis hybrids. Species distribution did not correlate with year of settlement nor with salinities taken at the time of sampling. There were significant differences in species distribution due to site location and depth, with M. edulis being more frequent at 8 m depth and M. trossulus more common at shallower depths. These differences might be exploitable in management strategies for continuing production, for example to decrease the prevalence of M. trossulus at shellfish farms in favour of the more desirable M. edulis.

Keywords

Blue musselHybridizationLochMe 15/16MytilusScotlandShellfish aquaculture

Introduction

While marine capture fisheries have not increased significantly in recent years and seem to have reached a maximum, with the proportion of overexploited and depleted stocks remaining stable over the last decade, aquaculture shows an average world annual growth rate of 8.8% per year since 1970, higher than any other animal food-producing sector (FAO 2006). This is equivalent to an increase in per capita supply from 0.7 kg in 1970 to 7.1 kg in 2004, accounting for 43% of the world total food fish supply (106 million t) in 2004 (FAO 2006).

Recent statistics indicate molluscs to be the third most important group of aquaculture species, in terms of both global production (13,243,000 t) and value (US$9,834 million) (FAO 2004, 2006). Within the molluscs, mussels are among the top 10 species groups produced in aquaculture, and their production has increased to 1,860,249 t in 2004 (FAO 2004, 2006).

Europe produces around 50% of the annual world production of mussels, consisting of Mytilus edulis from the Atlantic and North Sea coasts and M. galloprovincialis from the Atlantic coasts of Spain and Portugal and the Mediterranean Sea (Smaal 2002). Most of the current production in Europe originates from the historically big producers: Spain, France and the Netherlands, followed by a marked increase in production in the UK, Ireland and Norway (Smaal 2002).

Although modest in comparison to other European countries, shellfish production is a growth industry in Scotland, dominated by the blue mussel M. edulis. Shellfish production records are available from 1986, when a total of 262 tonnes of mussels was recorded. This increased rapidly in the following decade to 1,072 tonnes in 1996. The most recent survey indicates production of 4,287 tonnes from a total of 227 sites (FRS 1996, 2006).

In Scotland, mussels are mostly rope grown on long lines in sea lochs (fjordic inlets), and production depends exclusively on the settlement of natural seed in these lochs. Mussel aquaculture ropes are introduced to the lochs around February/March, at the beginning of the spawning season, and larval settlement occurs during the whole summer until September. Mussels are left to grow on the ropes until they reach an acceptable harvest size, between 2 and 3 years after settlement.

Mytilus edulis, M. galloprovincialis and M. trossulus are closely related species, whose identification, based on highly plastic morphological shell-shape characters, has long been controversial. More accurate species identification is now possible using genetic methods (Gosling 1992; Inoue et al. 1995), but even these are not always diagnostic (Wood et al. 2003). The three species are known to hybridise wherever their geographic ranges overlap, hybrids being fertile and able themselves to produce backcrosses (Beaumont et al. 1993, 2005; Wood et al. 2003). Well-studied hybrid zones include the area between southwest France and the Scottish coasts (M. edulis and M. galloprovincialis) (Gosling 1992; Hilbish et al. 2002; Skibinski et al. 1983), between the North Sea (M. edulis) and the Baltic Sea (M. trossulus) (Bierne et al. 2003; Riginos and Cunningham 2005) and on both the west (M. trossulus, M. galloprovincialis, M. edulis) and east coasts of North America where M. edulis and M. trossulus are found together (Braby and Somero 2006; Comesana et al. 1999; Johnson and Geller 2006, Rawson et al. 2001; Shields et al. 2008; Toro et al. 2004). No records exist of significant exploitation of M. trossulus in Europe, due mainly to the fragile shells, small size and low meat quality that seem to characterise this species in the Baltic Sea (Beaumont et al. 2007; Smaal 2002; Smietanka et al. 2004).

In 2004, growers reported the presence of fragile-shelled mussels in Loch Etive and consequent significant loss in production (around 25%), as the fragile-shelled mussels were destroyed during the harvest and grading processes (Beaumont et al. 2008). These fragile-shelled mussels in Loch Etive are characterised by an elongated “paddle-shaped” shell, with the dorsal and ventral edges being roughly parallel for some of their length. The shells are flexible and will gape when squeezed. In contrast, normal M. edulis-type mussel shells in Loch Etive are rigid and not elongated (Beaumont et al. 2008; Dias et al., observations in the present study). Beaumont et al. (2008) have demonstrated that fragile-shelled mussels from Loch Etive are morphologically and genetically distinct from M. edulis and were identified as being mostly M. trossulus or M. trossulus × M. edulis hybrids. In addition, small numbers of M. galloprovincialis and all possible hybrids between these three mussel species were identified in the loch. Fragile-shelled mussels were generally more frequent at shallower depths and closest to the riverine sources of the loch suggesting a salinity-related distribution (Beaumont et al. 2008).

The eastern Canadian mussel industry has been working for at least a decade on strategies for the production of M. edulis rather than M. trossulus, which is abundant at some cultivation sites. Research has addressed differences in shell strength (defined as the force causing shell breakage, which is of significance in processing), appearance (shell colour and shape, which affect consumer perception of quality), growth rates (measured based on shell width, depth, width-length ratio and depth-length ratio), production rates and survival between the two species (Canadian Aquaculture R&D Review 2007; Penney and Hart 1999; Penney et al. 2007).

This report builds on the initial detection of the three species of mussel in Loch Etive (Beaumont et al. 2008) by sampling at ten sites throughout the loch Etive in order to investigate their distribution in relation to key parameters such as depth, location and salinity. Mussels in the samples were individually identified to the “species” level based on PCR product fragment lengths at the Me15/16 locus (Inoue et al. 1995: M. edulis, 180 bp; M. trossulus, 168 bp; M. galloprovincialis, 126 bp). Scoring at this locus allows rapid and economical identification of each mussel genotype (Coghlan and Gosling 2007).

This study provides the industry and regulatory bodies with detailed information on the distribution of all mussel species and hybrids at aquaculture units in Loch Etive, and forms the basis of discussion of future management strategies aimed at limiting the impact of M. trossulus on mussel production in Scotland.

Materials and methods

Sampling

Mussels (1–5 years old) were collected from 10 aquaculture sites in Loch Etive, west of Scotland, at the end of January 2007 (Fig. 1). Site 1 was closest to the mouth of the loch and site 10 was farthest inland.
https://static-content.springer.com/image/art%3A10.1007%2Fs10499-008-9214-6/MediaObjects/10499_2008_9214_Fig1_HTML.gif
Fig. 1

Map (ArcGis©) showing sampling sites at Loch Etive (right image) in Scotland (left). Sites were numbered 1–10, site 1 being the closest to the Loch entrance and site 10 the most distant

One rope of mussels was randomly sampled at each site, and 30 adult mussels were taken haphazardly at each of 3 depths (2, 5, and 8 m from the surface) where they were available. Access to the sites was by boat, and mussels were collected by hand and stored in labelled net bags. The bags were transported in boxes with ice, prior to sample processing.

Salinity profiles were taken at each site at the time of sampling using a SAIV® CTD ST204 with Seapoint Fluorometer and turbidity meter. The exact locations of the sampling sites were determined using a hand-held GPS instrument.

DNA extraction, amplification and electrophoresis

Mussels were dissected and gill tissue sampled and preserved in 70% ethanol and stored at −20°C. DNA was extracted from approximately 0.5 mg of gill tissue from each mussel using a Qiagen BioRobot M48 and Qiagen M48 MagAttract DNA Mini Kit, following the manufacturer’s instructions. DNA (200 μl elution samples) was stored at −20°C.

PCR amplification was carried out using Me15/16 primers and conditions as described by Inoue et al. (1995), with an additional extension period of 5 min at 72°C following the final cycle. Each PCR was undertaken in a 25-μl reaction mixture containing 3 μl DNA, 5 mM dNTPs, 1 mM MgCl2, 0.4 mM of each primer, one unit of Taq DNA polymerase (Sigma) and PCR-grade water (Sigma). A negative control, with no template DNA added, was included in all PCR assays. PCR products from all sites were separated on 2% ethidium bromide-agarose gels alongside molecular weight markers and visualised under UV light.

Genetic analysis

Deviations from the Hardy–Weinberg expectations for the Me 15/16 locus in each sample were estimated from Fis values within FSTAT 2.9.3 (Goudet 1995). The significance of Fis was tested using permutation (1,000 replicates).

Statistical analysis

Distribution of genotype frequency over sampling sites and depths, and its possible relation with salinity and year of settlement was investigated using Generalized Linear Models in GenStat©.

Results

Site characteristics

Ten sites were sampled in January 2007 (Fig. 1). At one site, it was possible to take mussels from a rope that had been initially deployed in 2002 (site 7), but had not been harvested (at the request of other researchers who were performing analysis around this site). At two other sites, mussels were taken from ropes of 2006 settlement (sites 5 and 6), and at the remaining 7 sites from ropes of 2004 settlement. Sampling at sites 1, 4 and 5 was limited to 2 and 5 m, due to the shallow water at sites 1 and 4, and to the absence of settlement below 5 m at site 5.

Salinity generally increased with depth and was higher at sites closer to the entrance of the loch (Fig. 2). The salinity varied from around 2 practical salinity units (PSU) at 2 m depth for sites 8, 9 and 10, to around 12 PSU at 8 m depth at sites 1–3.
https://static-content.springer.com/image/art%3A10.1007%2Fs10499-008-9214-6/MediaObjects/10499_2008_9214_Fig2_HTML.gif
Fig. 2

Salinity (PSU) data measured at 10 sites and 3 depths (2, 5 and 8 m) from mussel rope farms in Loch Etive, Scotland, January 2007

Genotype frequencies

Scoring at the Me 15/16 locus was reliable and clear (only one individual, site 4, 2 m depth, failed to amplify) and revealed the presence of mussels from all three Mytilus species (M. trossulus, M. edulis and M. galloprovincialis) and their hybrids (Table 1). The M. galloprovincialis genotype was rare (only 4 individuals found in total) and was found at only three sites (site 3 at 8 m depth, site 5 at 5 m depth and site 6 at 5 and 8 m depth). M. galloprovincialis hybrids were more frequent and present, in small numbers, at all sites and depths. M. galloprovincialis × M. trossulus hybrids were found at all sites with an average proportion of 0.03, varying from 0.01 to 0.07, while M. galloprovincialis × M. edulis hybrids were present at all sites except site 9, with an average proportion of 0.07, varying from 0.0 to 0.14.
Table 1

Genotype proportions of Mytilus edulis (Me), M. trossulus (Mt), M. galloprovincialis (Mg) and their hybrids sampled at 10 sites and 3 depths (2, 5 and 8 m) from mussel rope farms in Loch Etive, Scotland

Species/hybrid

Site

1

2

3

4

5

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

Me

0.20

0.13

0.13

0.23

0.37

0.27

0.27

0.53

0.31

0.37

0.40

0.50

Mt

0.50

0.63

0.43

0.40

0.33

0.43

0.33

0.13

0.34

0.37

0.30

0.13

Mg

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.03

0.00

0.00

0.00

0.03

Me × Mt

0.23

0.17

0.40

0.30

0.20

0.23

0.37

0.13

0.14

0.20

0.13

0.13

Me × Mg

0.03

0.07

0.00

0.03

0.10

0.70

0.03

0.13

0.21

0.03

0.10

0.17

Mt × Mg

0.03

0.00

0.03

0.03

0.00

0.00

0.00

0.03

0.00

0.03

0.07

0.03

Total (n)

30

30

0

30

30

30

30

30

30

29

30

0

30

30

0

Species/hybrid

Site

6

7

8

9

10

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

Me

0.33

0.83

0.50

0.33

0.13

0.50

0.33

0.10

0.47

0.07

0.10

0.07

0.17

0.17

0.40

Mt

0.33

0.03

0.10

0.33

0.37

0.10

0.30

0.67

0.20

0.60

0.60

0.60

0.57

0.37

0.40

Mg

0.00

0.03

0.03

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Me × Mt

0.20

0.03

0.10

0.27

0.37

0.27

0.17

0.20

0.27

0.33

0.27

0.33

0.23

0.30

0.20

Me × Mg

0.13

0.07

0.23

0.00

0.07

0.07

0.13

0.03

0.00

0.00

0.00

0.00

0.03

0.07

0.00

Mt × Mg

0.00

0.00

0.03

0.07

0.07

0.07

0.07

0.00

0.07

0.00

0.03

0.00

0.00

0.10

0.00

Total (n)

30

30

30

30

30

30

30

30

30

30

30

30

30

30

30

M. trossulus and M. edulis genotypes and hybrids were found at all sites, proportions varying from 0.16 to 0.60, 0.08 to 0.56 and 0.11 to 0.31, respectively (Figs. 3, 4, 5). From the total individuals sampled, 30% were of M. edulis, 37% were M. trossulus and 23% were M. edulis × M. trossulus genotypes (Table 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10499-008-9214-6/MediaObjects/10499_2008_9214_Fig3_HTML.gif
Fig. 3

Proportion of the Mytilus edulis genotype at 10 sites, 3 depths (2, 5 and 8 m), and 3 years of settlement (2002, 2004, 2006) from mussel rope farms in Loch Etive, Scotland

https://static-content.springer.com/image/art%3A10.1007%2Fs10499-008-9214-6/MediaObjects/10499_2008_9214_Fig4_HTML.gif
Fig. 4

Proportion of the M. trossulus genotype at 10 sites, 3 depths (2, 5 and 8 m), and 3 years of settlement (2002, 2004, 2006) from mussel rope farms in Loch Etive, Scotland

https://static-content.springer.com/image/art%3A10.1007%2Fs10499-008-9214-6/MediaObjects/10499_2008_9214_Fig5_HTML.gif
Fig. 5

Proportion of the M. edulis × M. trossulus hybrid genotype at 10 sites, 3 depths (2, 5 and 8 m), and 3 years of settlement (2002, 2004, 2006) from mussel rope farms in Loch Etive, Scotland

Table 2

Numbers and proportions of M. edulis, M. trossulus and their hybrids sampled from 3 depths at 10 mussel rope farm sites in Loch Etive, Scotland

Species/hybrid

Number

Proportion

M. edulis

246

0.30

M. trossulus

297

0.37

M. edulis × M. trossulus

185

0.23

Others

81

0.10

Total

809

1.00

Frequencies of the three species specific alleles for all sites and depths, and globally across all sites are shown in Table 3. Genotype frequencies were in Hardy–Weinberg equilibrium at all depths and sites.
Table 3

Allele frequencies at the Me15/16 locus of mussels sampled from 10 sites and 3 depths on mussel rope farms in Loch Etive, Scotland

Species alleles

Site

1

2

3

4

5

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

Me

0.33

0.25

0.00

0.33

0.40

0.52

0.42

0.47

0.67

0.48

0.48

0.00

0.52

0.65

0.00

Mt

0.63

0.72

0.00

0.65

0.57

0.43

0.55

0.52

0.22

0.42

0.48

0.00

0.40

0.22

0.00

Mg

0.03

0.03

0.00

0.02

0.03

0.05

0.03

0.02

0.12

0.10

0.03

0.00

0.08

0.13

0.00

p of Fis

0.40

0.46

0.09

0.29

0.46

0.44

0.24

0.41

0.44

0.51

0.48

0.37

Species alleles

Site

6

7

8

9

10

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

2 m

5 m

8 m

Me

0.50

0.88

0.67

0.47

0.35

0.67

0.48

0.22

0.60

0.23

0.23

0.23

0.30

0.35

0.50

Mt

0.43

0.05

0.17

0.50

0.58

0.27

0.42

0.77

0.37

0.77

0.75

0.77

0.68

0.57

0.50

Mg

0.07

0.07

0.17

0.03

0.07

0.07

0.10

0.02

0.03

0.00

0.02

0.00

0.02

0.08

0.00

p of Fis

0.50

0.54

0.28

0.39

0.06

0.18

0.39

0.38

0.35

0.09

0.23

0.09

0.41

0.17

0.61

Me Allele 180 bp, M. edulis; Mt allele168 bp, M. trossulus; Mg allele126 bp M. galloprovincialis. p of Fis = probability of agreement with the Hardy–Weinberg model

Statistical analysis

The numbers of M. galloprovincialis and its hybrids were too small to carry out meaningful statistical analyses of their occurrence in relation to parameters.

Statistical analyses of the data for M. edulis, M. trossulus and their hybrids were carried out to assess the extent to which differences in the observed proportions of occurrence could be ascribed to site and depths effects. This was performed by fitting generalised linear models to the proportions assuming these to be binomially distributed. The estimated residual variances in the fitted models were found to be larger than for a binomial distribution, i.e. the data were overdispersed. The standard errors of the fitted site and depth effects from these models were therefore adjusted to take this overdispersion into account.

It was found that, in the model, both site and depth were significant (P < 0.05) in estimating the expected numbers of M. edulis, M. trossulus and their hybrids. The significant site effect is largely due to sites 5 and 6 having more M. edulis (45 and 58%, respectively), smaller numbers of M. trossulus (22 and 18%) and fewer hybrids (13 and 10%), and also to site 9 having few M. edulis (8%) and more M. trossulus (60%) and hybrids (30%). At a depth of 8 m, there are significantly more M. edulis than at depths of 2 and 5 m. Complementing this, there are fewer M. trossulus at 8 m but the numbers of hybrids are not significantly different from those found at other depths.

Further analyses were carried out to investigate whether the site and depth effects were related to year of settlement and salinity. Although sites 5 and 6 were both settled in 2006, the effect of year of settlement was not significant. Salinity was highest at 8 m depth at each site, but does not account for the variability in species distribution at shallower depths.

Discussion

In the present study, a total of 810 mussels were sampled from 10 sites, and mussel species and their hybrids were identified using Me 15/16 genotypes (Inoue et al. 1995). Based on this marker, the samples contained a high percentage of M. trossulus (37%) in cultivation that exceeded the prevalence of M. edulis (30%), and an average of 23% of the samples were M. trossulus × M. edulis hybrids.

Differences in the distribution of genotypes were significant when considering the influence of sites, and result largely from sites 5 and 6 having more M. edulis and smaller numbers of M. trossulus, and site 9 having fewer M. edulis and more M. trossulus. No consistent significant differences have been observed that could be related to site location, considering factors such as distance to the mouth of the loch (and hence seawater/freshwater flow influence), as suggested by Beaumont et al. (2008).

In this previous study, sampling was limited to two sites widely spaced in the loch. The much more comprehensive sampling in the present study, together with associated salinity data, show that neither site nor salinity alone appears to account for the differences in mussel species distribution in Loch Etive. However, because salinity was only estimated at a single point in time for all sites, these data do not provide information about either the average salinities, or the minima and maxima for the sites.

Differences in the distributions of M. trossulus and M. edulis on ropes were statistically significant at 8 m. Mytilus trossulus is more frequent and M. edulis less frequent in near-surface samples (2 and 5 m) than at 8 m depth. There were, however, no significant differences between depths of 2 and 5 m, nor between the distributions of M. trossulus × M. edulis hybrids with sampling depth.

Studies in Loch Etive in the 1990s did not report any unusual mussels (Okomus and Stirling 1998; Karayucel and Karayucel 2000). Mussels sampled from Loch Etive in 1999 were used as a reference population of presumed pure M. edulis by Wood et al. (2003), and the same genetic markers Me 15/16 identified all samples to be M. edulis homozygotes. Farmers reported significant numbers of fragile mussels in 2004 and, while recognising that they could have been present in smaller numbers previously, this suggests a dramatic increase of the M. trossulus population in just a few years. Genotype frequencies conformed to Hardy–Weinberg expectations at all sites and depths. Whilst such analysis is based upon a relatively small sample size, it suggests panmixia with no significant selection operating on hybrids and no pre- or post-mating reproductive barriers between the species.

Mytilus galloprovincialis and their hybrids were rare in cultivation in Loch Etive and their presence is not of known concern to the mussel industry in Scotland. Their presumed origin is the Mediterranean Sea and their presence on the northern coast of Scotland has been reported before (Skibinski et al. 1983). It is not known whether the northward expansion of their range in both the Atlantic and Pacific oceans is due to climate change or human activity but is probably a mixture of the two (Braby and Somero 2006; Coghlan and Gosling 2007; Wonham 2004; Zardi et al. 2007).

Differences in shell morphology (strength and appearance), survival, growth and production rates confer a clear commercial advantage to M. edulis as an aquaculture product compared with M. trossulus. Mixed species stocks have negative implications for the industry, and the presence of M. trossulus is recognized as being particularly problematic in Canada (Penney et al. 2002, 2006, 2007). Shell shape, for example, indirectly influences the buyer perception of overall product quality, affecting the farmers’ ability to market their product (Penney and Hart 1999; Penney et al. 2007). Differences in characteristics like shell shape between Mytilus species are not solely controlled by genotype, but also by a strong genotype × site effect (Shields et al. 2008), as shown by reciprocal mussel seed transfer experiments in the Atlantic Canadian hybrid zone (Penney et al. 2006, 2007).

Mussel aquaculture production in western Canada originated mainly (>75%) from the Prince Edward Island (PEI) area in the Gulf of St. Lawrence, and is based on unispecific local M. edulis stocks (Penney and Hart 1999). For many years, the potential for production was disregarded in the hybrid areas extending throughout most of the Canadian Atlantic coast from the Quebec north shore to Nova Scotia, where M. trossulus prevalence ranges from 9 to 60% (Penney and Hart 1999). In the last decade, however, expansion of shellfish aquaculture in areas outside the PEI has turned the attention of industry and governmental research bodies to this issue. They have been working together to develop and test management strategies with the objective of increasing the productivity of M. edulis/M. trossulus mixed species areas (Penney and Hart 1999, Penney et al. 2006).

Based on studies of species compositions, and given the genotypic inherent differences within populations in cultivation in Newfoundland, replacement of indigenous stocks in mixed species areas by unispecific M. edulis stocks is suggested as a strategy to enhance the industry production (Penney et al. 2002, 2006, 2007). Most recent research towards mussel cultivation has therefore focused on identifying good areas for seed collection (in terms of both quantity and quality), and on assessing the impact of seed transfer on species survival and growth (Canadian Aquaculture R&D Review 2007).

In Loch Etive, observations that M. edulis is more frequent below 5 m depth may present a useful initial approach to controlling the composition of mixed species stocks in cultivation. Spat collection at 8 m depth, for example, rather than in water near the surface would involve little effort and minor changes on the method of rope deployment during the seed settlement period only, and may prove to be a simple commercial way to increase the proportion of M. edulis over M. trossulus in farm stocks.

For seed substitution to be feasible in Loch Etive, it would be necessary to improve knowledge of Mytilus populations in Scotland, noting any other mixed species areas and identifying unispecific M. edulis areas where sufficient high quality seed might be produced for collection and transfer to other areas. Such a study would also be relevant to assessing the status of M. galloprovincialis in Scotland and any possible impacts on native populations.

Acknowledgements

The authors would like to thank Walter Spiers (Muckairn Mussels), Peter Richardson (Kaimes Fish Farming—Lorn Fisheries) and SI Sea farms for sampling support, providing the boat and mussels used in this study. Thank you also to Dave Fraser and the Association of Scottish Shellfish Growers for their support and advice, and to Steve Hay and David Tullet for providing and assisting with some of the tools used in this work. Joana Dias holds an ECOSUMMER PhD fellowship funded by the Marie Curie Research Training Network under the EU Sixth Framework Programme.

Copyright information

© Springer Science+Business Media B.V. 2008